Pricing
Simple pricing. Hard caps.
Start free with no card. Every plan has a hard usage cap, so you never pay more than the price on this page. Monthly, cancel in one click.
No card needed
Free
$0
~27M tokens / month
- The full agent, every language
- CLI and web app
- /workfor autonomous runs
- No API keys to manage
Most popular
Pro
$20/mo
~108M tokens / month
- Everything in Free
- 4× the Free allowance
- Unlimited /workfor runs
- Priority throughput
Heavy use
Max
$200/mo
~1.09B tokens / month
- Everything in Pro
- 10× the Pro allowance
- Highest throughput priority
- For agents running all day
For teams
Team, $1,000/mo flat
Up to 50 concurrent agents, unlimited seats, and an admin and usage dashboard. Start with a 30-day pilot ($1,500, credited toward your first two months).
How usage works
What the token numbers mean
Tokens measure how much text the model reads and writes. Long autonomous runs read more of your code, so they use more. Each plan's cap is set in dollars of usage, and most of what Sweet reads comes from cache, which costs less, so your allowance usually goes further than the token estimate. When you reach the cap, Sweet pauses until the next month or until you upgrade. There are no overage charges.
Billing questions
Can I cancel anytime?Yes. Plans are monthly. Cancel in one click from the billing page and you keep access until the end of the period you paid for.
Can I change plans?Yes. Upgrade from the billing page at any time. To move to a smaller plan, email hello@sweetcli.com and we'll switch it for you.
Who handles payment?Stripe. We never see or store your card details.
Do I need my own API key?No. Every plan includes model usage through Sweet, up to the cap.
Where does my code run?The CLI runs on your machine. The web app runs each session in its own isolated Firecracker microVM. Trust Center